I went ahead and added it as unixbench5 for now. Given the additional
requirements for v5, I'm also going to submit a patch to the v4 test
to fix the parallel make issues.

Unixbench5 patch here:
http://test.kernel.org/pipermail/autotest/2011-January/008135.html

- dale

On Fri, Jan 14, 2011 at 12:50 PM, Gregory P. Smith <[email protected]> wrote:
>
> On Fri, Jan 14, 2011 at 12:41 PM, Dale Curtis <[email protected]> wrote:
>>
>> After working with the unixbench team, I managed to get the parallel make
>> fixes pushed upstream. There's now a new version of unixbench available,
>> 5.1.3 with those fixes.
>> However, there's a hitch in that, as of version 5.0, the unixbench
>> launcher has been rewritten in perl. I wasn't able to find any other perl
>> scripts which were directly called by tests in the project, so I thought I
>> would ask here. What's the project policy on perl tests?
>> - dale
>
> well, when taking tests from somewhere and integrating them with runners
> into autotest you don't really get much choice.
> if perl is a problem on anyone's target test platform and they want to run
> this test, a suitable perl interpreter can be added to their client/deps.
> It might be nice to do the unixbench 5.x upgrade by creating a unixbench5
> client test instead of replacing the existing one.
> -gps
>>
>>
>> On Thu, Jan 13, 2011 at 11:36 AM, Gregory P. Smith <[email protected]> wrote:
>>>
>>> "yes please" on the update. :)
>>>
>>> On Thu, Jan 13, 2011 at 11:15 AM, Dale Curtis <[email protected]>
>>> wrote:
>>>>
>>>> Coming back to this, we've discovered unixbench is not parallel make
>>>> friendly. While debugging the issue, I noticed the 4.0.1 unixbench version
>>>> we're using in Autotest is pretty old; with the latest being 5.1.2. Before 
>>>> I
>>>> go about fixing the current version, I'm wondering if we should update to
>>>> the latest version. The latest version has supposedly been updated to 
>>>> handle
>>>> multi-CPU systems better.
>>>> http://code.google.com/p/byte-unixbench/
>>>>
>>>> Comments? Concerns?
>>>> - dale
>>>>
>>>> On Fri, Aug 13, 2010 at 8:36 PM, Amos Kong <[email protected]>
>>>> wrote:
>>>>>
>>>>> On Sat, Aug 14, 2010 at 4:53 AM, Eric Li(李咏竹) <[email protected]>
>>>>> wrote:
>>>>> > Thanks for the quick response. Please take another look.
>>>>> >
>>>>> > Index: client/common_lib/utils.py
>>>>> > ===================================================================
>>>>> > --- client/common_lib/utils.py (revision 4747)
>>>>> > +++ client/common_lib/utils.py (working copy)
>>>>> > @@ -1165,6 +1165,16 @@
>>>>> >      system('%s %s' % (configure, ' '.join(args)))
>>>>> >
>>>>> >
>>>>> > +def make(extra='', make='make', timeout=None, ignore_status=False):
>>>>> > +    """
>>>>> > +    Run make, adding MAKEOPTS to the list of options.
>>>>> > +
>>>>> > +    @param extra: extra command line arguments to pass to make.
>>>>> > +    """
>>>>> > +    cmd = '%s %s %s' % (make, os.environ.get('MAKEOPTS', ''), extra)
>>>>> > +    return system(cmd, timeout=timeout, ignore_status=ignore_status)
>>>>> > +
>>>>> > +
>>>>> >  def compare_versions(ver1, ver2):
>>>>> >      """Version number comparison between ver1 and ver2 strings.
>>>>>
>>>>> Looks good for me.
>>>>> ACK
>>>>> _______________________________________________
>>>>> Autotest mailing list
>>>>> [email protected]
>>>>> http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
>>>>
>>>>
>>>> _______________________________________________
>>>> Autotest mailing list
>>>> [email protected]
>>>> http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
>>>>
>>>
>>
>
>
_______________________________________________
Autotest mailing list
[email protected]
http://test.kernel.org/cgi-bin/mailman/listinfo/autotest

Reply via email to